CT accounts for 10% of radiological exams, but contributes -70% of the medical population dose. Little work has been done to determine doses to children. CT scans are currently not optimized, with scans performed at the same voltage (120 kV) irrespective of patient size and diagnostic task. We will measure doses and image quality when adult & pediatric patients undergo CT examinations. Experiments will be performed on six anthropomrphic phantoms (i.e., adult male & female, 10 yo; five yo; one yo and a newborn).Radiation dose measurements in single 2.5 cm thick slabs permit determination of energy imparted (e), effective dose (E) as functions of patient age and body region. Accurate measures of patient energy imparted will take into account the size and characteristics of the specific individual undergoing a CT examination. We will validate our approach to CT dosimetry by conducting field trials, which will also yield data on individual organ doses (and uncertainties) that could be used in future epidemilogical studies to investigate radiation risks. Our anthropomphic phantoms will permit measurement of signal to noise ratios (SNR) for important detection tasks in CT. Variations in SNR will be studied as a function of kVp, mAs and pitch ratio (PR). Three human observer models will be investigated to identify the one that best predicts observer performance. The human observer model will be applied to issues of inter-scanner variability and the optimal reconstruction filter for the detection of small low, contrast lesions. Dosimetery and image quality information will be combined to identify which CT parameters result in the lowest patient doses at equal imaging performance. We will also investigate the detection of simulated nodules in chest CT images to quantify the relative importance of structure/quantum in clinical CT examinations. Our image quality measurements will be performed in all body regions (head, chest, abdomen, pelvis) and patients who range from the newborn to the adult. These dose & image quality data will affect radiological practice in three ways: 1. Permit development of imaging protocols to ensure that patient doses are to kept as low as reasonably achievable (ALARA). 2. Determine the choice of kVp/mAs and PR/mAs that minimize patient doses with no loss of diagnostic performance. 3. Help the Radiology community to balance the confliciting requirements of image quality and patient doses in clinical CT.